Research map / Image generation / Autoregressive and latent generation
Image Disentanglement Autoencoder for Steganography Without Embedding
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use structure-texture disentanglement and image synthesis rather than conventional synthesis-based SWE | IDEAS · LSUN Bedrooms | FID: 147.45 -> 13.39 | Helped |
| Use structure-texture disentanglement and image synthesis rather than conventional synthesis-based SWE | IDEAS · LSUN Churches | FID: 99.59 -> 14.48 | Helped |
| Use structure-texture disentanglement and image synthesis rather than conventional synthesis-based SWE | IDEAS · FFHQ | FID: 67.95 -> 28.45 | Helped |
| Encode the secret in stable structure representations | IDEAS · LSUN Bedrooms, LSUN Churches, and FFHQ | extraction accuracy: 98.41% -> 100%; 97.53% -> 100%; 97.23% -> 100% | Helped |
| Increase N from 1 to 2 (ablation) | IDEAS · LSUN Bedrooms, σ = 1 | hidden capacity: 256 bits -> 512 bits | Helped |
| Increase ∆ from 0% to 50% (ablation) | IDEAS · LSUN Bedrooms, σ = 1 | extraction accuracy: 100% -> 99.54% | Hurt |
| Increase ∆ from 0% to 50% (ablation) | IDEAS · LSUN Bedrooms, σ = 1 | FID: 16.88 -> 13.39 | Helped |
| Increase ∆ from 0% to 50% (ablation) | IDEAS · LSUN Churches, σ = 1 | extraction accuracy: 100% -> 99.55% | Hurt |
| Increase ∆ from 0% to 50% (ablation) | IDEAS · FFHQ, σ = 1, N = 1 | extraction accuracy: 100% -> 99.49% | Hurt |
| Increase σ from 1 to 3 (ablation) | IDEAS · LSUN Bedrooms, N = 2, ∆ = 0% | extraction accuracy: 100% -> 76.34% | Hurt |
| Increase σ from 1 to 3 (ablation) | IDEAS · LSUN Churches, N = 2, ∆ = 0% | extraction accuracy: 100% -> 79.60% | Hurt |
| Increase σ from 1 to 3 (ablation) | IDEAS · FFHQ, N = 2, ∆ = 0% | extraction accuracy: 100% -> 81.23% | Hurt |
Closest papers by meaning
- StegFormer: Rebuilding the Glory of Autoencoder-Based Steganography (AAAI 2024 · similarity 0.87)
- SSHR: More Secure Generative Steganography with High-Quality Revealed Secret Images (ICML 2025 · similarity 0.85)
- StegaStyleGAN: Towards Generic and Practical Generative Image Steganography (AAAI 2024 · similarity 0.85)
- Robust Message Embedding via Attention Flow-Based Steganography (CVPR 2025 · similarity 0.85)
- Provably Secure Image Robust Steganography via Cross-modal Error Correction (AAAI 2025 · similarity 0.84)
- StegoGAN: Leveraging Steganography for Non-Bijective Image-to-Image Translation (CVPR 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.